"AI is writing 90% of our code" sounds impressive before you realize that AI-generated code is orders of magnitude more verbose & less efficient than code written by a professional software engineer.
But "we ship 9 lines of fluff for each line of code that does something" doesn't sound as impressive.
@PavelASamsonov If the vibecoded implementations were reliably correct - but they are not; while trying to understand better what I criticise, I am trying out GenAI in software engineering in isolated use cases and environments, but the failure modes range from hilarious to insidious.
Many still follow these output-related (quantity) metrics rather than outcome-related ones (quality).
@PavelASamsonov Some people are still stuck on the old "lines of code per day" metric. Which has never been a decent metric anyway, but is less and less so. If, as a dev, I manage 10 lines of new code - written, tested, unit tested, designed, considered - in a day, that is pretty reasonable. Sometimes, my work is negative lines in a day.
What they mean is that 90% of their code is now unmaintainable.
Curious Carrot
Pavel A. Samsonov
Greg Lloyd
Monika Barget
Michael Hartle
Sven A. Schmidt
Schroedingers_Cat